[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort

haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort URL: https://github.com/apache/incubator-mxnet/pull/15501#discussion_r301889632 ## File path: python/mxnet/numpy/multiarray.py ## @@ -1670,6 +1675,60 @@ def argmax(a, axis=None,

[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort

haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort URL: https://github.com/apache/incubator-mxnet/pull/15501#discussion_r301889586 ## File path: python/mxnet/symbol/numpy/_symbol.py ## @@ -379,13 +379,58 @@ def topk(self, *args,

[GitHub] [incubator-mxnet] haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort

haojin2 commented on a change in pull request #15501: [Numpy] Numpy compatible argsort URL: https://github.com/apache/incubator-mxnet/pull/15501#discussion_r301889649 ## File path: python/mxnet/ndarray/numpy/_op.py ## @@ -425,6 +426,60 @@ def argmax(a, axis=None,

[GitHub] [incubator-mxnet] iblis17 commented on issue #15454: Julia docs

iblis17 commented on issue #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#issuecomment-509896736 BTW, Julia v1.0.4 is out. Let's test it: https://github.com/apache/incubator-mxnet/pull/15502.

[GitHub] [incubator-mxnet] iblis17 opened a new pull request #15502: CI: upgrade Julia version from 1.0.3 to 1.0.4

iblis17 opened a new pull request #15502: CI: upgrade Julia version from 1.0.3 to 1.0.4 URL: https://github.com/apache/incubator-mxnet/pull/15502 This is a bugfix release. No breaking changes. see: https://julialang.org/downloads/

[incubator-mxnet] branch ib/bump-jl10 created (now 907f066)

This is an automated email from the ASF dual-hosted git repository. iblis pushed a change to branch ib/bump-jl10 in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git. at 907f066 CI: upgrade Julia version from 1.0.3 to 1.0.4 This branch includes the following new

[incubator-mxnet] 01/01: CI: upgrade Julia version from 1.0.3 to 1.0.4

This is an automated email from the ASF dual-hosted git repository. iblis pushed a commit to branch ib/bump-jl10 in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git commit 907f0d700e10535c57ea8f1d0d7a288da446 Author: Iblis Lin AuthorDate: Wed Jul 10 11:48:37 2019 +0800

[GitHub] [incubator-mxnet] iblis17 commented on a change in pull request #15454: Julia docs

iblis17 commented on a change in pull request #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301875253 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,93 @@ Refer to the [Clojure setup

[GitHub] [incubator-mxnet] iblis17 commented on a change in pull request #15454: Julia docs

iblis17 commented on a change in pull request #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301875253 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,93 @@ Refer to the [Clojure setup

[GitHub] [incubator-mxnet] anirudhacharya commented on issue #14836: Refactor AGInfo and Imperative

anirudhacharya commented on issue #14836: Refactor AGInfo and Imperative URL: https://github.com/apache/incubator-mxnet/pull/14836#issuecomment-509893815 @larroy why does this PR include changes to the tvm submodule? This is

[GitHub] [incubator-mxnet] reminisce commented on issue #13143: [MXNET-1206] Support NDArray indexing with None and Ellipsis

reminisce commented on issue #13143: [MXNET-1206] Support NDArray indexing with None and Ellipsis URL: https://github.com/apache/incubator-mxnet/pull/13143#issuecomment-509893086 @kohr-h Gentle ping. Any updates on the performance tests? This PR is useful in getting NumPy consistent

[GitHub] [incubator-mxnet] iblis17 commented on a change in pull request #15454: Julia docs

iblis17 commented on a change in pull request #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301869452 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,93 @@ Refer to the [Clojure setup

[GitHub] [incubator-mxnet] iblis17 commented on a change in pull request #15454: Julia docs

iblis17 commented on a change in pull request #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301868736 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,93 @@ Refer to the [Clojure setup

[GitHub] [incubator-mxnet] Plusmonkey commented on issue #15473: mac mxnet cpu compile error

Plusmonkey commented on issue #15473: mac mxnet cpu compile error URL: https://github.com/apache/incubator-mxnet/issues/15473#issuecomment-509885806 @lanking520 @pengzhao-intel Hi, I update the summary and hot to reproduce it.

[GitHub] [incubator-mxnet] wkcn commented on issue #14894: Accelerate ROIPooling layer

wkcn commented on issue #14894: Accelerate ROIPooling layer URL: https://github.com/apache/incubator-mxnet/pull/14894#issuecomment-509883779 @sxjscience @KellenSunderland Could you please help take a review? Thank you!

[GitHub] [incubator-mxnet] wkcn commented on issue #15323: Two fixes for info_gan.md example Code

wkcn commented on issue #15323: Two fixes for info_gan.md example Code URL: https://github.com/apache/incubator-mxnet/pull/15323#issuecomment-509883421 Merged. Thank you: ) This is an automated message from the Apache Git

[incubator-mxnet] branch master updated: Two fixes for info_gan.md example Code (#15323)

This is an automated email from the ASF dual-hosted git repository. wkcn pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 7d4d1bc Two fixes for info_gan.md

[GitHub] [incubator-mxnet] wkcn merged pull request #15323: Two fixes for info_gan.md example Code

wkcn merged pull request #15323: Two fixes for info_gan.md example Code URL: https://github.com/apache/incubator-mxnet/pull/15323 This is an automated message from the Apache Git Service. To respond to the message, please

[incubator-mxnet] branch master updated: Opperf: Support Python<3.6 (#15487)

This is an automated email from the ASF dual-hosted git repository. wkcn pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new d82c89a Opperf: Support Python<3.6

[GitHub] [incubator-mxnet] wkcn commented on issue #15487: Opperf: Support Python<3.6

wkcn commented on issue #15487: Opperf: Support Python<3.6 URL: https://github.com/apache/incubator-mxnet/pull/15487#issuecomment-509882322 Merged. Thanks for your contribution! This is an automated message from the Apache

[GitHub] [incubator-mxnet] wkcn merged pull request #15487: Opperf: Support Python<3.6

wkcn merged pull request #15487: Opperf: Support Python<3.6 URL: https://github.com/apache/incubator-mxnet/pull/15487 This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [incubator-mxnet] wkcn commented on issue #14031: Fix transposed convolution in CPU w/o MKLDNN.

wkcn commented on issue #14031: Fix transposed convolution in CPU w/o MKLDNN. URL: https://github.com/apache/incubator-mxnet/pull/14031#issuecomment-509880566 @apeforest Hi! Any update in this PR? The PR is important: ) This

[GitHub] [incubator-mxnet] mikemwx opened a new pull request #15501: [Numpy] Numpy compatible argsort

mikemwx opened a new pull request #15501: [Numpy] Numpy compatible argsort URL: https://github.com/apache/incubator-mxnet/pull/15501 ## Description ## Numpy compatible argsort with existing kernals. [Numpy argsort

[GitHub] [incubator-mxnet] mikemwx commented on issue #15277: [Numpy] Numpy argsort

mikemwx commented on issue #15277: [Numpy] Numpy argsort URL: https://github.com/apache/incubator-mxnet/pull/15277#issuecomment-509877599 Because I accidentally used merge when starting this development branch, this branch is now closed and a new pull request is made with clear and clean

[GitHub] [incubator-mxnet] mikemwx closed pull request #15277: [Numpy] Numpy argsort

mikemwx closed pull request #15277: [Numpy] Numpy argsort URL: https://github.com/apache/incubator-mxnet/pull/15277 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub

[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.

This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new f005213 Bump the publish

[GitHub] [incubator-mxnet] lanking520 opened a new pull request #15500: fix the bug on Scala Sparse

lanking520 opened a new pull request #15500: fix the bug on Scala Sparse URL: https://github.com/apache/incubator-mxnet/pull/15500 ## Description ## This is a fix of the tests failure on Scala CPU. I also enable the full integration test coverage on CPU. @zachgk @nswamy ##

[GitHub] [incubator-mxnet] ZhennanQin commented on issue #15465: [RFC] Integrate TVM into Apache MXNet

ZhennanQin commented on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509862734 @yidawang If we all agree that we should avoid using more than one thread pool implementation in the runtime, then I guess OMP is

[incubator-mxnet] branch master updated: [Perl] - simplify aliasing strategy (#15395)

This is an automated email from the ASF dual-hosted git repository. sergeykolychev pushed a commit to branch master in repository https://gitbox.apache.org/repos/asf/incubator-mxnet.git The following commit(s) were added to refs/heads/master by this push: new 6b2b927 [Perl] - simplify

[GitHub] [incubator-mxnet] sergeykolychev merged pull request #15395: [Perl] - simplify aliasing strategy

sergeykolychev merged pull request #15395: [Perl] - simplify aliasing strategy URL: https://github.com/apache/incubator-mxnet/pull/15395 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [incubator-mxnet] yidawang commented on issue #15465: [RFC] Integrate TVM into Apache MXNet

yidawang commented on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509861413 @ZhennanQin I guess we should first make sure that we don't use more than one thread pool implementation in the runtime. Also,

[GitHub] [incubator-mxnet] ZhennanQin commented on issue #15465: [RFC] Integrate TVM into Apache MXNet

ZhennanQin commented on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509859385 @tqchen I agree with you that this is a technical issue and shouldn't block this PFC. As thread management should be considered

[GitHub] [incubator-mxnet] DickJC123 commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage.

DickJC123 commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage. URL: https://github.com/apache/incubator-mxnet/pull/15449#discussion_r301844223 ## File path: src/common/cuda_utils.cc ## @@ -0,0 +1,116 @@ +/* + * Licensed to

[GitHub] [incubator-mxnet] larroy commented on issue #15285: Graph dumper

larroy commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509858209 @ptrendx Thanks This is an automated message from the Apache Git Service. To respond to the

[GitHub] [incubator-mxnet] larroy commented on issue #15270: Fix warnings in CLang.

larroy commented on issue #15270: Fix warnings in CLang. URL: https://github.com/apache/incubator-mxnet/pull/15270#issuecomment-509857187 @roywei sure. This is an automated message from the Apache Git Service. To respond to

[GitHub] [incubator-mxnet] larroy commented on issue #14836: Refactor AGInfo and Imperative

larroy commented on issue #14836: Refactor AGInfo and Imperative URL: https://github.com/apache/incubator-mxnet/pull/14836#issuecomment-509854327 Is this good to go? This is an automated message from the Apache Git Service.

[GitHub] [incubator-mxnet] larroy opened a new pull request #15499: Improve diagnose.py, adding build features info and binary library path.

larroy opened a new pull request #15499: Improve diagnose.py, adding build features info and binary library path. URL: https://github.com/apache/incubator-mxnet/pull/15499 See title. This is the script used to fill issues, so adding more info will help. Refined output:

[GitHub] [incubator-mxnet] aaronmarkham commented on issue #15454: Julia docs

aaronmarkham commented on issue #15454: Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#issuecomment-509852422 > Could you paste the preview link? It's in the main description: http://34.201.8.176/versions/julia/api/julia/site/index.html But since the

[GitHub] [incubator-mxnet] larroy commented on issue #15424: fixed config.mk and Makefile bugs for installing mkl

larroy commented on issue #15424: fixed config.mk and Makefile bugs for installing mkl URL: https://github.com/apache/incubator-mxnet/pull/15424#issuecomment-509847294 Indeed this is a mess and we should improve it. I suggest having this PR merged if anyone doesn't have any additional

[GitHub] [incubator-mxnet] larroy edited a comment on issue #15424: fixed config.mk and Makefile bugs for installing mkl

larroy edited a comment on issue #15424: fixed config.mk and Makefile bugs for installing mkl URL: https://github.com/apache/incubator-mxnet/pull/15424#issuecomment-509847294 Indeed this is messy and we should improve it, is difficult to reason about and track these flags. I suggest

[GitHub] [incubator-mxnet] larroy edited a comment on issue #15424: fixed config.mk and Makefile bugs for installing mkl

larroy edited a comment on issue #15424: fixed config.mk and Makefile bugs for installing mkl URL: https://github.com/apache/incubator-mxnet/pull/15424#issuecomment-509847294 Indeed this is a mess and we should improve it, is difficult to reason and track these flags. I suggest having

[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU

anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#discussion_r301798181 ## File path: docs/faq/env_var.md ## @@ -309,6 +309,17 @@ If ctypes is used, it must be

[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU

anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#discussion_r301822032 ## File path: tests/python/gpu/test_fusion.py ## @@ -0,0 +1,195 @@ +# Licensed to the Apache Software

[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU

anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#discussion_r301820981 ## File path: src/operator/fusion/fused_op.cc ## @@ -0,0 +1,269 @@ +/* + * Licensed to the Apache Software

[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU

anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#discussion_r301815807 ## File path: src/executor/pointwise_fusion_pass.cc ## @@ -0,0 +1,275 @@ +/* + * Licensed to the Apache

[GitHub] [incubator-mxnet] anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU

anirudh2290 commented on a change in pull request #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#discussion_r301822614 ## File path: src/operator/fusion/fused_op.cc ## @@ -0,0 +1,269 @@ +/* + * Licensed to the Apache Software

[GitHub] [incubator-mxnet] adis300 closed pull request #15301: Ignore generated nnvm.cc template. This file is created whenever `make` is run

adis300 closed pull request #15301: Ignore generated nnvm.cc template. This file is created whenever `make` is run URL: https://github.com/apache/incubator-mxnet/pull/15301 This is an automated message from the Apache Git

[GitHub] [incubator-mxnet] adis300 opened a new pull request #15301: Ignore generated nnvm.cc template. This file is created whenever `make` is run

adis300 opened a new pull request #15301: Ignore generated nnvm.cc template. This file is created whenever `make` is run URL: https://github.com/apache/incubator-mxnet/pull/15301 Ignore generated nnvm.cc template in amalgamation dir.

[GitHub] [incubator-mxnet] adis300 closed pull request #15488: Deprecate m_hard LDFLAGS.

adis300 closed pull request #15488: Deprecate m_hard LDFLAGS. URL: https://github.com/apache/incubator-mxnet/pull/15488 This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [incubator-mxnet] adis300 opened a new pull request #15488: Deprecate m_hard LDFLAGS.

adis300 opened a new pull request #15488: Deprecate m_hard LDFLAGS. URL: https://github.com/apache/incubator-mxnet/pull/15488 Forgot to remove the flags in PR #15435. The previous PR fixes the issue but this PR makes it cleaner. mhard-float option has been deprecated by Google.

[GitHub] [incubator-mxnet] ArmageddonKnight commented on issue #14383: MXNET_BACKWARD_DO_MIRROR is broken

ArmageddonKnight commented on issue #14383: MXNET_BACKWARD_DO_MIRROR is broken URL: https://github.com/apache/incubator-mxnet/issues/14383#issuecomment-509839196 My implementation is available here:

[GitHub] [incubator-mxnet] ArmageddonKnight commented on issue #14383: MXNET_BACKWARD_DO_MIRROR is broken

ArmageddonKnight commented on issue #14383: MXNET_BACKWARD_DO_MIRROR is broken URL: https://github.com/apache/incubator-mxnet/issues/14383#issuecomment-509838577 @antinucleon Given below is the outputs that I got from the LSTM-based NMT model (from the *Sockeye* toolkit): ###

[GitHub] [incubator-mxnet] ptrendx edited a comment on issue #15285: Graph dumper

ptrendx edited a comment on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509835138 > For feedback to be useful it has also to be concrete, in this case I don't see the feedback to my code change proposal is concrete enough for me to take

[GitHub] [incubator-mxnet] ptrendx commented on issue #15285: Graph dumper

ptrendx commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509835138 > For feedback to be useful it has also to be concrete, in this case I don't see the feedback to my code change proposal is concrete enough for me to take

[GitHub] [incubator-mxnet] larroy commented on issue #14836: Refactor AGInfo and Imperative

larroy commented on issue #14836: Refactor AGInfo and Imperative URL: https://github.com/apache/incubator-mxnet/pull/14836#issuecomment-509825978 @mxnet-label-bot update [Backend, pr-awaiting-review] This is an automated

[GitHub] [incubator-mxnet] larroy commented on issue #14836: Refactor AGInfo and Imperative

larroy commented on issue #14836: Refactor AGInfo and Imperative URL: https://github.com/apache/incubator-mxnet/pull/14836#issuecomment-509825640 @mxnet-label-bot update [Backend, pr-awaiting-review] This is an automated

[GitHub] [incubator-mxnet] larroy opened a new pull request #15498: Add -DMXNET_USE_OPENMP to Makefiles so runtime info gets updated accordingly

larroy opened a new pull request #15498: Add -DMXNET_USE_OPENMP to Makefiles so runtime info gets updated accordingly URL: https://github.com/apache/incubator-mxnet/pull/15498 ## Description ## Fixes OPENMP info in runtime features.

[GitHub] [incubator-mxnet] larroy commented on issue #15285: Graph dumper

larroy commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509815270 @ptrendx not getting defensive, that was not my intention, let's not read between the lines. Feedback should be concrete an actionable for efficient use of

[GitHub] [incubator-mxnet] ptrendx commented on issue #15455: Improve docs for AMP

ptrendx commented on issue #15455: Improve docs for AMP URL: https://github.com/apache/incubator-mxnet/pull/15455#issuecomment-509809028 Number of commits seems wrong - you should probably rebase this PR before merging. Otherwise, LGTM.

[GitHub] [incubator-mxnet] ptrendx commented on issue #15167: Pointwise fusion for GPU

ptrendx commented on issue #15167: Pointwise fusion for GPU URL: https://github.com/apache/incubator-mxnet/pull/15167#issuecomment-509805767 @szha @KellenSunderland @eric-haibin-lin Do you have any comments to this PR? This

[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.

This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new 29fed3d Bump the publish

[GitHub] [incubator-mxnet] ptrendx commented on issue #15285: Graph dumper

ptrendx commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509795602 Also, to clarify - I do not think that reusing JSON dump and other utilities is a good approach to this. My concern is only about the new graph class itself

[GitHub] [incubator-mxnet] ptrendx commented on issue #15285: Graph dumper

ptrendx commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509790305 @larroy There is no need to get so defensive about this, I (and I assume others as well) just want to make your contribution better as a result of the review

[GitHub] [incubator-mxnet] KhurramPirov edited a comment on issue #15479: Clear optimizer state in batch.end.callback

KhurramPirov edited a comment on issue #15479: Clear optimizer state in batch.end.callback URL: https://github.com/apache/incubator-mxnet/issues/15479#issuecomment-509532475 @lanking520, @Linhaibin this is the step where I want to clear optimizer state, I want to make momentums of

[GitHub] [incubator-mxnet] larroy edited a comment on issue #15331: [fix] missing input log higher order.

larroy edited a comment on issue #15331: [fix] missing input log higher order. URL: https://github.com/apache/incubator-mxnet/pull/15331#issuecomment-509778127 I'm still not sure what's the meaning of the backward output for the head gradient input as we discussed before. This week we are

[GitHub] [incubator-mxnet] larroy edited a comment on issue #15331: [fix] missing input log higher order.

larroy edited a comment on issue #15331: [fix] missing input log higher order. URL: https://github.com/apache/incubator-mxnet/pull/15331#issuecomment-509778127 I'm still not sure what's the meaning of the backward output for the head gradient input as we discussed before. This week we are

[GitHub] [incubator-mxnet] larroy commented on issue #15331: [fix] missing input log higher order.

larroy commented on issue #15331: [fix] missing input log higher order. URL: https://github.com/apache/incubator-mxnet/pull/15331#issuecomment-509778127 I'm still not sure what's the meaning of the backward output for the head gradient input as we discussed before. This week we are at a

[GitHub] [incubator-mxnet] larroy commented on issue #15285: Graph dumper

larroy commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509777153 @ptrendx I didn't see code that dumps directly to dot, but to Json as discussed above. If you guys have such a problem with introducing a graph class which has

[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage.

larroy commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage. URL: https://github.com/apache/incubator-mxnet/pull/15449#discussion_r301756893 ## File path: src/common/cuda_utils.cc ## @@ -0,0 +1,116 @@ +/* + * Licensed to the

[GitHub] [incubator-mxnet] larroy commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage.

larroy commented on a change in pull request #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage. URL: https://github.com/apache/incubator-mxnet/pull/15449#discussion_r301756667 ## File path: src/common/cuda_utils.cc ## @@ -0,0 +1,116 @@ +/* + * Licensed to the

[GitHub] [incubator-mxnet] aaronmarkham commented on a change in pull request #15454: [WIP] Julia docs

aaronmarkham commented on a change in pull request #15454: [WIP] Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301751246 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,86 @@ Refer to the [Clojure setup

[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.

This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new fceb394 Bump the publish

[GitHub] [incubator-mxnet] aaronmarkham commented on a change in pull request #15454: [WIP] Julia docs

aaronmarkham commented on a change in pull request #15454: [WIP] Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#discussion_r301750481 ## File path: docs/install/ubuntu_setup.md ## @@ -310,25 +310,86 @@ Refer to the [Clojure setup

[GitHub] [incubator-mxnet] keerthanvasist commented on issue #15496: Loaded pretrained model but train accuracy starts from zero,

keerthanvasist commented on issue #15496: Loaded pretrained model but train accuracy starts from zero, URL: https://github.com/apache/incubator-mxnet/issues/15496#issuecomment-509754676 @mxnet-label-bot add [python, question]

[GitHub] [incubator-mxnet] keerthanvasist commented on issue #15492: No CMAKE_CUDA_COMPILER could be found

keerthanvasist commented on issue #15492: No CMAKE_CUDA_COMPILER could be found URL: https://github.com/apache/incubator-mxnet/issues/15492#issuecomment-509753903 @mxnet-label-bot add [build, cmake, cuda] This is an

[GitHub] [incubator-mxnet] roywei commented on issue #15431: [MXNet 1.5.0.rc2] Issues with asnumpy() method

roywei commented on issue #15431: [MXNet 1.5.0.rc2] Issues with asnumpy() method URL: https://github.com/apache/incubator-mxnet/issues/15431#issuecomment-509750539 Hi @Wallart are you able to come up with a reproduciable example? I suspect it may only occur when using dataloader with

[GitHub] [incubator-mxnet] yugoren commented on issue #15424: fixed config.mk and Makefile bugs for installing mkl

yugoren commented on issue #15424: fixed config.mk and Makefile bugs for installing mkl URL: https://github.com/apache/incubator-mxnet/pull/15424#issuecomment-509742449 > @yugoren Please find the definition of `MSHADOW_USE_MKL` at

[GitHub] [incubator-mxnet] aaronmarkham commented on issue #15454: [WIP] Julia docs

aaronmarkham commented on issue #15454: [WIP] Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#issuecomment-509742169 @iblis17 I tried to add some more details on getting Ubuntu configured natively (since all I did here was get it working via Docker and to run the

[GitHub] [incubator-mxnet] DickJC123 commented on issue #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage.

DickJC123 commented on issue #15449: cuda/cuDNN lib version checking. Force cuDNN v7 usage. URL: https://github.com/apache/incubator-mxnet/pull/15449#issuecomment-509741064 Not sure if the timing permits this, but I'd think this might be a useful PR to backport to v1.5

[GitHub] [incubator-mxnet] yidawang commented on issue #15465: [RFC] Integrate TVM into Apache MXNet

yidawang commented on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509730974 We will never want to use two thread pools simultaneously. We should use the one which gives best-to-date performance. Technically

[GitHub] [incubator-mxnet] tqchen edited a comment on issue #15465: [RFC] Integrate TVM into Apache MXNet

tqchen edited a comment on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509712700 Technically it is easy to drop in openmp as a threadpool backend of TVM. It is just not the default one. Historically tvm

[GitHub] [incubator-mxnet] ptrendx commented on issue #15285: Graph dumper

ptrendx commented on issue #15285: Graph dumper URL: https://github.com/apache/incubator-mxnet/pull/15285#issuecomment-509722904 I also do not understand why you need a whole new Graph class for the purpose of serializing to dot. nnvm::Graph and nnvm::IndexedGraph should give you

[GitHub] [incubator-mxnet] aaronmarkham commented on issue #15454: [WIP] Julia docs

aaronmarkham commented on issue #15454: [WIP] Julia docs URL: https://github.com/apache/incubator-mxnet/pull/15454#issuecomment-509716984 > @aaronmarkham It isn't true anymore. Thanks, I'll change that and rebase so maybe this CI/maven problem will be fixed. How far should

[GitHub] [incubator-mxnet] tqchen commented on issue #15465: [RFC] Integrate TVM into Apache MXNet

tqchen commented on issue #15465: [RFC] Integrate TVM into Apache MXNet URL: https://github.com/apache/incubator-mxnet/issues/15465#issuecomment-509712700 Re: Threadpool Technically it is easy to drop in openmp as a threadpool backend of TVM. It is just not the default one.

[GitHub] [incubator-mxnet] adis300 closed pull request #15488: Deprecate m_hard LDFLAGS.

adis300 closed pull request #15488: Deprecate m_hard LDFLAGS. URL: https://github.com/apache/incubator-mxnet/pull/15488 This is an automated message from the Apache Git Service. To respond to the message, please log on to

[GitHub] [incubator-mxnet] zixuanweeei edited a comment on issue #15497: Independent gradients requests check with respect to weights and bias of convolution

zixuanweeei edited a comment on issue #15497: Independent gradients requests check with respect to weights and bias of convolution URL: https://github.com/apache/incubator-mxnet/pull/15497#issuecomment-509675440 > Could you try to add a UT for this case? Sure. The existent UT passed

[GitHub] [incubator-mxnet] zixuanweeei commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution

zixuanweeei commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution URL: https://github.com/apache/incubator-mxnet/pull/15497#issuecomment-509675440 > Could you try to add a UT for this case? Sure. The exists UT passed with

[GitHub] [incubator-mxnet] pengzhao-intel commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution

pengzhao-intel commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution URL: https://github.com/apache/incubator-mxnet/pull/15497#issuecomment-509671515 Could you try to add a UT for this case?

[GitHub] [incubator-mxnet] zixuanweeei commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution

zixuanweeei commented on issue #15497: Independent gradients requests check with respect to weights and bias of convolution URL: https://github.com/apache/incubator-mxnet/pull/15497#issuecomment-509667543 @pengzhao-intel @ciyongch @TaoLv Please help me review on this PR. Thanks. 

[GitHub] [incubator-mxnet] francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn

francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn URL: https://github.com/apache/incubator-mxnet/pull/15495#discussion_r301618570 ## File path:

[GitHub] [incubator-mxnet] francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn

francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn URL: https://github.com/apache/incubator-mxnet/pull/15495#discussion_r301618858 ## File path:

[GitHub] [incubator-mxnet] francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn

francis0407 commented on a change in pull request #15495: [Numpy] Added operator logaddexp; added support for zero-size tensor in BinaryBroadcastBackwardUseIn URL: https://github.com/apache/incubator-mxnet/pull/15495#discussion_r301618570 ## File path:

[GitHub] [incubator-mxnet] zixuanweeei opened a new pull request #15497: Independent gradients requests check with respect to weights and bias of convolution

zixuanweeei opened a new pull request #15497: Independent gradients requests check with respect to weights and bias of convolution URL: https://github.com/apache/incubator-mxnet/pull/15497 ## Description ## As it was described in #15464, MXNet with MKL-DNN gives a wrong gradient of a

[incubator-mxnet-site] branch asf-site updated: Bump the publish timestamp.

This is an automated email from the ASF dual-hosted git repository. marcoabreu pushed a commit to branch asf-site in repository https://gitbox.apache.org/repos/asf/incubator-mxnet-site.git The following commit(s) were added to refs/heads/asf-site by this push: new a0f940b Bump the publish

[GitHub] [incubator-mxnet] cloud-oak opened a new pull request #15323: Two fixes for info_gan.md example Code

cloud-oak opened a new pull request #15323: Two fixes for info_gan.md example Code URL: https://github.com/apache/incubator-mxnet/pull/15323 ## Description ## Fix the InfoGan Tutorial. ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your

[GitHub] [incubator-mxnet] cloud-oak closed pull request #15323: Two fixes for info_gan.md example Code

cloud-oak closed pull request #15323: Two fixes for info_gan.md example Code URL: https://github.com/apache/incubator-mxnet/pull/15323 This is an automated message from the Apache Git Service. To respond to the message,

[GitHub] [incubator-mxnet] braindotai commented on issue #15429: Operator Performance Regression on CPU

braindotai commented on issue #15429: Operator Performance Regression on CPU URL: https://github.com/apache/incubator-mxnet/issues/15429#issuecomment-509612532 According to the above mentioned benchmark results, it looks like v1.5 is gonna be more faster than v1.4. Am I right?

[GitHub] [incubator-mxnet] avivna edited a comment on issue #15486: mxnet profiler yields exception when multithreaded mxndarray IO operations occur in the background

avivna edited a comment on issue #15486: mxnet profiler yields exception when multithreaded mxndarray IO operations occur in the background URL: https://github.com/apache/incubator-mxnet/issues/15486#issuecomment-509598878 In order to make recreation of this bug easier, I have recreated

[GitHub] [incubator-mxnet] KhurramPirov opened a new issue #15496: Loaded pretrained model but train accuracy starts from zero,

KhurramPirov opened a new issue #15496: Loaded pretrained model but train accuracy starts from zero, URL: https://github.com/apache/incubator-mxnet/issues/15496 ## Description I loaded pretrained model, but got zero train accuracy (acc) at the same time pretrained

[GitHub] [incubator-mxnet] Maicus commented on issue #15484: Binding Model fails with simple_bind error

Maicus commented on issue #15484: Binding Model fails with simple_bind error URL: https://github.com/apache/incubator-mxnet/issues/15484#issuecomment-509599536 So i got it to work by adding the Target and the TargetCode Section as well. But the size of these should be dependent of

  1   2   >